Measuring discharge pressure of centrifugal pump

[shantanu@india]

Hi,
Centrifugal pumps generate head by throwing water through the viens of the rotating impeller & this invariably creates a pressure pulsation in the discharge line.
Now what is the most accurate way to measure this pressure....a bourdon tube type pressure gauge or a piezoelectric type pressure transducer?
Conventional wisdom obviously favours the electronic method since it is supposed to be the more modern & updated method.But yesterday I was proved wrong!!
I had connected a Emerson 0-25Kg/cm2 pressure transmitter generating 4-20mA & a Wika glycerine filled bourdon gauge to the same point of the pump discharge.
The pressure trasnmitter was consistently showing 0.1/0.2Kg higher pressure than the pressure gauge & the readings were inconsistent with the true readings .
I have surmised that this is happening due to the pressure pulsation...the low inertia peizo responds very fast to pressure fluctuations & the filtering algorithm generates a value which is higher(& erroneous) than the mechanically damped reading of the pressure gauge.
Now my question is what should be the right correction algorithm to  be applied to the raw electronic pressure signal? Should root mean sqare yield better results than mean? Should the algorithm take into account the RPM for fine tuning?

It would not be out of place to mention here that calibrating both the pressure gauge & the pressure transmitter by a Dead Weight Tester have provided accurate results....most probably due to the fact that a Dead Weight Testes generates hydrostatic pressure whereas a centrifugal pump generates dynamic pressure pulses.

[TimB]

I myself am having issues with pulsation on  a pressure sensor due to cavitation.

This though is not a new problem and the solution seems to be, to use a pulse arrestor. This turns out to be either something like a rubber diaphragm or a needle valve so the pulse has to squeeze through a tight hole

Search for "pressure gauge damper"

[gtvpic]

The installation of pressure gauges and transmitters is highly recommended to have a hydrostatic filter duct, or simply a piece of rolled tube that minimizes the effects of centrifugal pumps.
Simply with a threaded tube, the results of the measurements are greatly improved with a real-time cost of updating them.

[shantanu@india]

Thanks Tim and gtvpic.
I'll explore the ideas

[joesaliba]

Some years ago we had some pumps that would start and stop with a mechanical switch. No inverter was used. The problem was that upon start there was a spike in pressure, which stopped the pump, followed by a drop of pressure and the pump restarted, and continued like that. ON--OFF--ON--OFF....and so on.

The solution was to make a small but efficient pulse arrestor to the pressure switch.

Practically, a small round piece metal drill with a 1.5mm drill to restrict water flow. You can see the difference with a pressure gauge, going up slowly.

Regards

Joe

[shantanu@india]

Separate snubbers were put in line with the gauge and the transmitter which significantly improved performance after adjustment. Some electronic fine tuning would further close the gap. Thanks everyone.